论文部分内容阅读
t检验是W. S. Gosset在1908年提出的一种适合于小样本资料的统计检验方法,由于检验效能高,计算相对简便,在科研工作中得到广泛应用,也为医学科研工作者所熟悉。但是,在一些医学论文中有时会发现错用、误用t检验的情况,究其原因,主要是对t检验的应用范围和适用条件还不够清楚。下面针对在医学论文中较为常见的几个t检验应用不当问题,进行一下分析讨论。 1 未考虑到资料的分布特征,对于呈明显偏态分布的数据也用t检验进行处理
The t-test is a statistical test method proposed by W. S. Gosset in 1908, which is suitable for small sample data. Because of its high test efficiency and relatively simple calculation, it has been widely used in scientific research and is also familiar to medical researchers. However, in some medical papers, cases of misuse or misuse of the t-test are sometimes found. The reason for this is mainly because the application range and applicable conditions of the t-test are still not clear enough. The following is a discussion of the inappropriate application of several t-tests that are common in medical papers. 1 The distribution characteristics of the data were not taken into account, and the data with a clearly skewed distribution was also processed using the t-test.